Re: How to determine database market share?

"hpuxrac" <johnbhurley_at_sbcglobal.net> wrote:

> Paul what you have posted doesn't make a lot of sense.

OK, a cogent argument - you may not agree with my initial premise, but at least I'm prepared to listen to you (might have something to do with your meaningful contributions to to .oracle lists over the time I've been reading them).  

> Something about measuring database market penetration by how many job
> ads there are?

<cease mutual back-slapping>

It was not "Something about measuring database market penetration by how many job ads there are?" - particular emphasis on the word "something".

I am genuinely interested in trying to get past the "market-babble" of various vendors and establish *_some_* sort of <scientific/reasonable/reliable/realistic/honest> way of determining market share of various db's in various markets.

Surveys - zealots who vote 100's of times/Consultants who tell the client what they want to hear.

SEC filings - try getting any meaningful info about number of working licenced installs from those numbers - what with ancilliary services, VAR agreements, corporate agreements, service agreements... the list goes on and on...  

> On the DB2 side, many of the people who are DB2 DBA's have been doing
> it for a relatively long time and don't often switch jobs. So that
> would impact any chance for anything meaningful for what you are
> talking about.

Oh," puhlease"... what you are saying may be true in a snapshot of a given industry at a given time, but if there's a huge pool of DB2 installs out there, then there will be a certain number of job advertisements for such positions.

DBA's die, they get run over by busses, they cross train, they go to Thailand and start to raise crocodiles for leather - whatever the hell happens in the lives of other people also happens (on average) to DB2 DBA's. Stating that their particular product and/or mode-de-vie is inherently stable (with respect to, say Oracle DBA's) doesn't hold a lot of water with me.  

> That's really pretty similar on the oracle side except there is
> somewhat more movement from company to company of people with Oracle
> DBA job skills.

As per above, you seem to be arguing that there is something inherently different about DB2 DBA's as opposed to Oracle ones. I'd be very interested in learning about this previously unknown nirvana that IBM DB2 DBA's have achieved - as opposed to their Oracle counterparts.

<smart arse>

do they smoke lots of weed?

</smart arse>.

I would argue that *_ON AVERAGE_* an Oracle DBA is as likely as an IBM DB2 DBA to change jobs because of <partner getting new job in different city, going to raise goats in the West of Ireland, .... insert reason of your choice>.

What is it, in your opinion that makes DB2 professionals that much more stable than Oracle ones?  

> On the SQL Server side, there are 2 questions involved, how many real
> SQL Server DBA jobs are there out there as opposed to developers who
> also support SQL Server, and how often does a company train/promote
> internal candidates into the SQL Server role.

Maybe as many as there are DB2 developers and/or Oracle developers who are *_ALSO_* moved sideways and upwards into a DBA role?

Basically, you are hand-waving and waffling! I want to be able to put a *_FACT_DRIVEN_* number beside a given DB and say that "Here in Ireland we have ..... " , "but in the Czech Republic, things are slightly differernt because.... "

> Certainly there are
> times when a company does actually hire from outside to support SQL
> Server but as a general rule I don't think so.

Where are you from? With a name like "johnbhurley", I'd assume some Irish ancestry - I went to school with a guy called John Hurley.

Anyway, my point is that *_WHEREVER THE HELL YOU ARE FROM_*, that *_ON AVERAGE_* over a longish period of time, job adverts are about the best metric you can get for db penetration in a marketplace.

Sure, some companies will move people around - "the new programmer guy, he's good with db's - we'll make him the dba - save us hiring somebody new". All *_PERFECTLY LOGICAL_*.

What is also perfectly logical, is that

1. not every company is going to do that

and

b) following from a - companies are going to have requirements from time to time to hire dba's - yes?

My argument is that from a very simple statistical viewpoint is that *_ON AVERAGE_* the number of job ads is going to reflect *_ON AVERAGE_* the number of deployments of a particular db (Oracle, IBM, SQL Server, Informix, MySQL.........), particularly in a county like America (look at dice.com) where the number of jobs on any given day is huge (10's of thousands) and therefore unlikely to suffer from extreme fluctuations more likely in small numbers.

I fail to see how anyone could fault that logic.  

> I guess you could attempt to make some kind of case to be able to
> measure "market demand" "when hiring from outside" but I don't think
> you could go any further than that.

<Driving point home>

"when hiring from outside" *_OVERALL_* and *_IN THE LONG TERM_* and *_ON AVERAGE_* is likely to be a reflection of what *_OVERALL_* market (i.e. company) demands are. This will *_BY ITS VERY NATURE_* reflect the market as a whole and hence provide an insight into what products are being used by what percentage of the *_OVERALL_* marketplace.

If you are incapable of grasping the ideas I have outlined above, I suggest you stick to
alt.moderate.intelligence.discussion.bullshitters.

Paul...

-- 

plinehan __at__ yahoo __dot__ __com__

XP Pro, SP 2, 

Oracle, 9.2.0.1.0 (Enterprise Ed.)
Interbase 6.0.1.0;

When asking database related questions, please give other posters 
some clues, like operating system, version of db being used and DDL.
The exact text and/or number of error messages is useful (!= "it didn't work!").
Thanks.
 
Furthermore, as a courtesy to those who spend 
time analysing and attempting to help, please 
do not top post.